Routine Catalog Research Articles

Performance of AI-Based Phase Picking and Event Association Methods after the Large 2023 Mw 7.8 and 7.6 Türkiye Doublet

ABSTRACT On 6 February 2023, a devastating earthquake doublet consisting of Mw 7.8 and 7.6 events separated by about 9 hr struck the southeastern part of Türkiye. The developing aftershock sequence contained thousands of events during the first few days and overwhelmed the routine algorithms handling their detection and location. In addition, several stations temporarily lost real-time contact and came online again later. At the same time the Omori decay of the aftershock event rate reduced the event frequency and allowed for inclusion of progressively smaller-magnitude events with time. One possibility to help deal with such a complex situation is the use of machine learning (ML) methods to generate earthquake catalogs with a substantially higher number of events. Here, we present high-resolution earthquake catalogs derived with two ML association methods for the first five days of the aftershock sequence of this doublet. In terms of the number of reliably located events, the event catalog created from PhaseNet picks and the GENIE phase association method outperforms both the routine regional catalog and the second ML-derived catalog obtained from the GaMMA phase association method. Although both GaMMA and GENIE catalogs detect about 6 times more events than the routine catalog, GENIE associates on average about double the phases to a single event than GaMMA, which results in better constrained event locations. The spatiotemporal evolution of the event rates is sensitive to changes in the network geometry due to variable station availability. During the first few days, no decay of the event rate in the enhanced catalog is observed due to the inclusion of progressively smaller-magnitude events with time and increased station availability. This study indicates that ML-derived earthquake catalogs for challenging time periods like the early aftershock sequences of large earthquakes have the potential to significantly improve routine event catalogs.

The seismicity in the middle section of the Altyn Tagh Fault system revealed by a dense nodal seismic array

The left-lateral Altyn Tagh Fault (ATF) system is the northern boundary of the Qinghai-Xizang Plateau, separating the Tarim Basin and the Qaidam Basin. The middle section of ATF has not recorded any large earthquakes since 1598 AD, so the potential seismic hazard is unclear. We develope an earthquake catalog using continuous waveform data recorded by the Tarim-Altyn-Qaidam dense nodal seismic array from September 17 to November 23, 2021 in the middle section of ATF. With the machine learning-based picker, phase association, location, match and locate workflow, we detecte 233 earthquakes with ML -1-3, far more than 6 earthquakes in the routine catalog. Combining with focal mechanism solutions and the local fault structure, we find that seismic events are clustered along the ATF with strike-slip focal mechanisms and on the southern secondary faults with thrusting focal mechanisms. This overall seismic activity in the middle section of the ATF might be due to the northeastward transpressional motion of the Qinghai-Xizang Plateau block at the western margin of the Qaidam Basin.

High-resolution seismicity imaging and early aftershock migration of the 2023 Kahramanmaraş (SE Türkiye) MW7.9 & 7.8 earthquake doublet

We build a high-resolution early aftershock catalog for the 2023 SE Türkiye seismic sequence with PALM, a seamless workflow that sequentially performs phase picking, association, location, and matched filter for continuous data. The catalog contains 29,519 well-located events in the two mainshocks rupture region during 2023-02-01–2023-02-28, which significantly improves the detection completeness and relocation precision compared to the public routine catalog. Employing the new PALM catalog, we analyze the structure of the seismogenic fault system. We find that the Eastern Anatolian Fault (EAF) that generated the first MW7.9 mainshock is overall near-vertical, whereas complexities are revealed in a small-scale, such as subparallel subfaults, unmapped branches, and stepovers. The seismicity on EAF is shallow (<15 km) and concentrated in depth distribution, indicating a clear lock-creep transition. In contrast, the Sürgü Fault (SF) that is responsible for the second MW7.8 mainshock is shovel-shaped for the nucleation segment and has overall low dip angles (∼40°–80°). Aftershocks on the SF distribute in a broad range of depth, extending down to ∼35 km. We also analyze the temporal behavior of seismicity, discovering no immediate foreshocks within ∼5 days preceding the first mainshock, and no seismic activity on the SF before the second mainshock.

Deep Learning-Based Microseismic Detection and Location Reveal the Seismic Characteristics and Causes in the Xiluodu Reservoir, China

ABSTRACT The Xiluodu reservoir, as the third reservoir developed in the lower Jinsha River, is the fourth largest reservoir in the world in terms of power generation. It is located in an area of historically high seismic intensity. A large amount of seismic activity has occurred in the reservoir area because the reservoir was impounded in 2013, but the mechanism of seismogenesis is still not clear. In this study, we collected continuous seismic records from July 2020 to October 2022 in the Xiluodu reservoir area, built a high-precision microseismic catalog for this region based on a deep learning seismic detection and location workflow called LOC-FLOW, and eventually obtained high-precision locations of 4924 earthquakes (five times more than the routine catalog). We sketched the main seismogenic structures based on the spatial and temporal distribution of the earthquakes in the catalog. According to the relationship between periodic variation of water level and seismic activity, seismicity in the reservoir area is active at the stage when the water level is filling to the highest point and starts to draw down. Especially, the sudden change in the rate of water level variation can easily trigger seismic activity. Combined with the spatiotemporal distribution of seismicity in each region and the previous results of numerical simulation, we concluded that the seismic activity in the reservoir head area and around the Manao fault is likely induced by the increase of normal stress and pore pressure diffusion caused by reservoir impoundment, whereas the ML 4.6 earthquake that occurred at the intersection of the Lianfeng fault and the Zhongcun fault was likely tectonic activity occurring on a concealed fault.

Along-Strike Variation of Seismicity Near the Extinct Mid-Ocean Ridge Subducted Beneath the Manila Trench

AbstractThe change in seismic activity is explored using data recorded by ocean-bottom seismometers (OBSs) and permanent seismic stations near the extinct Mid-Ocean ridge of the South China Sea (SCS) and the Manila trench. We apply the machine learning–based algorithm EQTransformer to the OBS dataset for seismic event detection and phase picking and then evaluate the precision and compare the time residuals between automatic and manual picks. We derive a catalog of earthquakes in the region and find bending-fault earthquakes in the outer rise at the northern of the Huangyan (Scarborough) Seamount chain, where no historical seismicity was reported in the routine catalog. Abundant outer-rise earthquakes occurred on both sides of the Huangyan (Scarborough) Seamounts chain, but the focal depths vary along the trench. The Wadati–Benioff zone of the eastward subducted SCS oceanic lithosphere can be clearly identified. The focal depths are down to ∼100 km near Luzon island at ∼16° N but deepen southward to a depth of ∼180 km at ∼14° N. Dips of the slab also steepen from north to south, indicating along-strike changes in the geometry of the Manila megathrust.

Spatiotemporal Aftershock Evolution of the 2014 M 6.4 and 5.9 Kangding Double Shocks in Sichuan, Southwestern China

Abstract We apply the graphics processing unit-based match and locate method to detect missing small aftershocks of the 2014 M 6.4 and 5.9 Kangding double shocks (hereafter KD1 and KD2, respectively) in Sichuan, southwestern China. KD1 and KD2 occurred on two adjacent branch faults of the Xianshuihe fault zone, respectively. We use the relocated routine catalog events as templates to scan through the continuous waveforms from KD1 to approximately 5 days after KD2. Approximately 7.8 times more aftershocks than the routine catalog are obtained. Our results show that the aftershocks occurring between the two mainshocks first expanded northwestward along the KD1 seismogenic fault and then migrated southeastward to the KD2 seismogenic fault. This feature may reflect the stress transfer process after KD1, which may represent the triggering process of KD1 to KD2. These aftershocks also show minor migration with logarithmic time in the along-strike and up-dip directions, which suggests that the afterslip may occur between the two mainshocks and improve the stress level around KD2 to some extent. Based on the improved aftershock catalog, we also calculate the b-values within different space–time windows in the aftershock zone. A relatively low b-value around the KD2 nucleation point after KD1 and before KD2 is observed, which may indicate that this area contained a high stress level and might be in the critical state of rupture, in which KD2 occurred at this particular moment. Our study suggests that after a moderate-to-large earthquake a more complete aftershock catalog can be obtained by detecting microearthquakes with which the triggering processes of subsequent mainshocks could be revealed; based on the detailed spatiotemporal distribution of aftershocks and b-values, the subsequent moderate-to-large earthquake hazard could be assessed to some extent.

Rapid Characterization of the July 2019 Ridgecrest, California, Earthquake Sequence From Raw Seismic Data Using Machine‐Learning Phase Picker

AbstractThe two principle earthquakes of the July 2019 Ridgecrest, California, earthquake sequence, MW 6.4 and 7.1, and their immediate foreshocks and thousands of aftershocks present a challenging environment for rapid analysis and characterization of this sequence as it unfolded. In this study, we analyze the first 6 days of the sequence using continuous data from available seismic networks to detect and locate earthquakes associated with the earthquake sequence. We build a high‐precision earthquake catalog using a deep‐neural‐network‐based picker—PhaseNet and a sequential earthquake association and location workflow. Without prior information, we automatically detect and locate more than twice as many earthquakes as the routine catalog. Our high‐precision earthquake catalog reveals detailed spatiotemporal evolution of the earthquake sequence and clearly defines multiple faults activated during the sequence. Our study demonstrates that it is possible to characterize earthquake sequences from raw seismic data using a well‐trained machine‐learning picker and our workflow.

Graphics Processing Unit-Based Match and Locate (GPU-M&amp;L): An Improved Match and Locate Method and Its Application

AbstractMicroearthquake detection and location are critical for understanding earthquake mechanisms and mitigating seismic hazards. Match and locate (M&amp;L) is an effective method for simultaneously detecting and locating small earthquakes. However, the heavy computational demands of the M&amp;L make it challenging to apply to big data. In this article, we develop an improved M&amp;L method—called graphics processing unit-based M&amp;L (GPU-M&amp;L). The GPU-M&amp;L differs from the M&amp;L in two ways: (1) adding weighting factor for each component of templates to improve the detection ability and (2) implementing the M&amp;L method on GPU to accelerate the computation. Synthetic tests show the GPU-M&amp;L can not only handle smaller earthquakes than the M&amp;L but also perform 4.5 times faster than the M&amp;L parallelly programed on central processing unit. As an example, we utilize the GPU-M&amp;L to study the seismic activity during seven days after the 2015 Ms 5.8 Alxa, China, earthquake (from 15 to 21 April 2015). Using 38 cataloged earthquakes as templates, we detect ∼20 times more events than in the routine catalog. The distribution of those detected events, along with focal mechanisms of large events, suggests that the 2015 Ms 5.8 earthquake occurred on an east–west-trending hidden strike-slip fault.

Rapid Earthquake Association and Location

ABSTRACTRapid association of seismic phases and event location are crucial for real‐time seismic monitoring. We propose a new method, named rapid earthquake association and location (REAL), for associating seismic phases and locating seismic events rapidly, simultaneously, and automatically. REAL combines the advantages of both pick‐based and waveform‐based detection and location methods. It associates arrivals of different seismic phases and locates seismic events primarily through counting the number of P and S picks and secondarily from travel‐time residuals. A group of picks are associated with a particular earthquake if there are enough picks within the theoretical travel‐time windows. The location is determined to be at the grid point with the most picks, and if multiple locations have the same maximum number of picks, the grid point among them with smallest travel‐time residuals. We refine seismic locations using a least‐squares location method (VELEST) and a high‐precision relative location method (hypoDD). REAL can be used for rapid seismic characterization due to its computational efficiency. As an example application, we apply REAL to earthquakes in the 2016 central Apennines, Italy, earthquake sequence occurring during a five‐day period in October 2016, midway in time between the two largest earthquakes. We associate and locate more than three times as many events (3341) as are in Italy's National Institute of Geophysics and Volcanology routine catalog (862). The spatial distribution of these relocated earthquakes shows a similar but more concentrated pattern relative to the cataloged events. Our study demonstrates that it is possible to characterize seismicity automatically and quickly using REAL and seismic picks.

Spatial-temporal distribution of early aftershocks following the 2016 Ms 6.4 Menyuan, Qinghai, China Earthquake

The Lenglongling fault, located in the north Qilian fold zone along the northeastern margin of the Tibetan Plateau, plays an important role in accommodating the tectonic deformation between the Tibetan Plateau and the Alxa block. On 21 January 2016, the Ms 6.4 Menyuan earthquake occurred near the Lenglongling fault. In this study, we utilize the waveform matched filter technique and double-difference earthquake location algorithm to examine the spatial-temporal evolution of aftershocks following the Menyuan mainshock. Events from both China Earthquake Networks Center (CENC) routine catalog and relocated catalog are used as templates to scan through the continuous waveform data (21 January 2016 to 31 December 2016). We identify three and eleven times more events than listed in CENC routine catalog and relocated catalog, respectively. We find that the earliest aftershocks (e.g., in the first 10 min or so) mostly illuminated a steeply dipping plane, while aftershocks at later times outlined a near vertical dipping plane with mostly strike-slip focal mechanisms. The relocated aftershocks also suggest that the Menyuan mainshock occurred in a secondary fault rather than the major Lenglongling fault, which is consistent with geodetic and geological observations. Moreover, we find that aftershocks systematically expand with time along the fault strike and 16 repeating earthquake clusters occur in the rupture zone, which are likely driven by afterslip of the mainshock.

Fluid‐faulting evolution in high definition: Connecting fault structure and frequency‐magnitude variations during the 2014 Long Valley Caldera, California, earthquake swarm

AbstractAn extended earthquake swarm occurred beneath southeastern Long Valley Caldera between May and November 2014, culminating in three magnitude 3.5 earthquakes and 1145 cataloged events on 26 September alone. The swarm produced the most prolific seismicity in the caldera since a major unrest episode in 1997–1998. To gain insight into the physics controlling swarm evolution, we used large‐scale cross correlation between waveforms of cataloged earthquakes and continuous data, producing precise locations for 8494 events, more than 2.5 times the routine catalog. We also estimated magnitudes for 18,634 events (~5.5 times the routine catalog), using a principal component fit to measure waveform amplitudes relative to cataloged events. This expanded and relocated catalog reveals multiple episodes of pronounced hypocenter expansion and migration on a collection of neighboring faults. Given the rapid migration and alignment of hypocenters on narrow faults, we infer that activity was initiated and sustained by an evolving fluid pressure transient with a low‐viscosity fluid, likely composed primarily of water and CO2exsolved from underlying magma. Although both updip and downdip migration were observed within the swarm, downdip activity ceased shortly after activation, while updip activity persisted for weeks at moderate levels. Strongly migrating, single‐fault episodes within the larger swarm exhibited a higher proportion of larger earthquakes (lower Gutenberg‐Richterbvalue), which may have been facilitated by fluid pressure confined in two dimensions within the fault zone. In contrast, the later swarm activity occurred on an increasingly diffuse collection of smaller faults, with a much higherbvalue.

Identifying Active Structures Using Double-Difference Earthquake Relocations in Southwest British Columbia and the San Juan Islands, Washington

This paper applies double-difference earthquake relocation techniques to investigate sources of seismicity in southwest British Columbia, Canada, and the San Juan Islands, Washington. The study area is a complex region of deformation and has the potential for large earthquakes in the North Americancrust. Double-difference earthquake relocation techniques are applied to identify otherwise-hidden active struc- tures that may pose a hazard to nearby population and infrastructure. We present evi- dence for previously unrecognized active structures using precise relative earthquake relocations obtained using both catalog and waveform cross-correlation data. Results have significantly reduced errors over routine catalog locations and show lineations in areas of clustered seismicity. In southwest British Columbia, these lineations or streaks appear to be hidden structures that do not disrupt near-surface sediments; however, in theSanJuanIslandstheidentifiedlineationcouldberelatedtorecentlymappedsurface expressions offaults identifiedfromseismic reflection andmultibeam bathymetric sur- veys.We use avariety of velocity models for therelocations and find that inappropriate models lead to artifacts at layer boundaries and increased vertical errors.

Empirical Relationship between Moment Magnitude and Nuttli Magnitude for Small-magnitude Earthquakes in Southeastern Canada

Other| November 01, 2005 Empirical Relationship between Moment Magnitude and Nuttli Magnitude for Small-magnitude Earthquakes in Southeastern Canada Eleanor Sonley; Eleanor Sonley *Now at Boston University Search for other works by this author on: GSW Google Scholar Gail M. Atkinson Gail M. Atkinson Search for other works by this author on: GSW Google Scholar Author and Article Information Eleanor Sonley *Now at Boston University Gail M. Atkinson Publisher: Seismological Society of America First Online: 09 Mar 2017 Online ISSN: 1938-2057 Print ISSN: 0895-0695 © 2005 by the Seismological Society of America Seismological Research Letters (2005) 76 (6): 752–755. https://doi.org/10.1785/gssrl.76.6.752 Article history First Online: 09 Mar 2017 Cite View This Citation Add to Citation Manager Share Icon Share Facebook Twitter LinkedIn MailTo Tools Icon Tools Get Permissions Search Site Citation Eleanor Sonley, Gail M. Atkinson; Empirical Relationship between Moment Magnitude and Nuttli Magnitude for Small-magnitude Earthquakes in Southeastern Canada. Seismological Research Letters 2005;; 76 (6): 752–755. doi: https://doi.org/10.1785/gssrl.76.6.752 Download citation file: Ris (Zotero) Refmanager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentBy SocietySeismological Research Letters Search Advanced Search The most common catalog magnitude for eastern North America is the Nuttli magnitude (Nuttli, 1973), referred to as MN or MbLg. The Nuttli magnitude is based on the amplitude of the Lg phase (multiply reflected and refracted shear waves). MN is quickly and easily determined from the time series, making it convenient for routine catalog purposes. The moment magnitude scale (Hanks and Kanamori, 1979) is generally preferred by seismologists for many applications, as it is more reflective of the actual size of the earthquake. Moment magnitude (M), however, is more... You do not have access to this content, please speak to your institutional administrator if you feel you should have access.

Relocation of the 26 July 2001 Skyros Island (Greece) earthquake sequence using the double-difference technique

Routine catalog phase data from the permanent stations of the Greek National Seismological Network were used to relocate 445 aftershocks of the 26 July 2001 M 6.5 Skyros (North Aegean Sea) earthquake, using the double-difference (DD) algorithm of Waldhauser and Ellsworth [Bull. Seism. Soc. Am. 19 (2000) 1353]. The relocated epicenters define two zones: (a) a main NW–SB cluster of ∼27 km length and (b) a secondary less pronounced NE–SW cluster of ∼15 km length, which was activated within 24 h after the mainshock occurrence. Redistribution of stresses due to the occurrence of the mainshock certainly played a role in this activation. The Skyros mainshock occurred in a region controlled by the activity of the Aegean Sea strands of the dextral strike-slip North Anatolian Fault (NAF). The main aftershock zone is almost perpendicular to the main tectonic NAF strands, and confirms that the causative fault strikes NW–SE associated with sinistral strike-slip motion. The 2001 Skyros earthquake is the first instrumentally recorded event that indicates re-activation of the old tectonic NW–SE trending structures of continental Greece under the presently acting stress field in a way sketched by Kiratzi [Geophys. J. Int. 151 (2002) 360].